Forecasting Outward Foreign Direct Investment by China Using Markov-Chain-Based Grey Residual Modification Model

Peiyi Kong ； Yi-Chung Hu

《Journal of Management & Decision Sciences》 2卷1期 (2019/06) Pp. 1-7

https://doi.org/10.6885/JMDS.201906_2(1).0001

摘要

Outward foreign direct investment (OFDI) from China has increased each year, especially since the implementation of the Belt and Road strategy in 2015. Under this circumstance, it is important to forecast the variability in China's OFDI. In this paper, annul data from 2011 to 2017 were collected to forecast the China's OFDI by establishing the proposed Markov-chain-based grey residual modification (MC-RGM(1,1)) model. In order to clarify the prediction accuracy of the proposed model, this paper compared the prediction results with GM(1,1) model and residual modification GM(1,1) model. In addition, the optimal input subset method is employed to decide the best length of the dataset. The results show that the prediction accuracy of MC-RGM(1,1) is superior than that of other prediction models for model-fitting and for ex post testing, which proved the validity of proposed MC-RGM(1,1) model. Understanding the trend of OFDI provides great support for Chinese government to consider the OFDI strategy. The results showed that the mean absolute percentage errors (MAPE) for the original GM(1,1), residual modification GM(1,1), and Markov-chain-based residual modification GM(1,1) models for model-fitting were 0.66%, 0.25%, and 0.31%, respectively, and the results of ex post testing the MAPE were 1.03%, 0.31%, and 0.28%, respectively. The experiment results verified that the proposed prediction models performed well compared with the original model.

關鍵字

Outward foreign direct investment ； Forecasting ； GM(1,1) ； Residual ； Markov chain